464 RETURN TO THE MlRAN SITE [Chap.XH 003 (three pieces); iv. ooii; vi. cox; ix. 002 (twenty pieces), 003 (twenty-four pieces), 004; x. 006; xill ooi, c.; xiv. 0074; xv-xvi 004; xxiv. 0040 (eleven pieces). It is, therefore, all the more noteworthy that among all these scales there is not one which; in shape or method of lacing, agrees with the type represented by the Niya Site specimen of mail. Whereas in this type the scales have a rounded upper end and are laced through three holes on each of the longer sides and one at the top end, the scales of the Miran fort are all of oblong shape and have different and, in some ways, more complicated systems of lacing. Fortunately, these can be studied in some sets where the scales are still joined to their neighbours by the original fastening of narrow leather thongs, as seen in the reproductions of Plate L (M. i. ix. 003; xxiv. 0040). That the scales belonged to a number of different suits of armour and varied in size as well as omamenta- jn ornamentation will be clearly seen on reference to the details recorded in. the several entries of sc°a!e& the Descriptive List Yet, at the same time, we note a certain uniformity in style and technical make as well as in the methods of lacing. The pieces of hard ' green' leather, perhaps of camel skin, vary considerably in size, from an average of af" x 2" to about 4^" x ^—' (c£ M. I. ix. 002, 003), but are always oblong with a slight outward curving. Originally all of them bore thick lustrous lacquer on both sides, generally applied in successive coats, the number of which varies up to seven (M. i, xxiv. 0040). Among the colours used, a brilliant red and black are by far the most prevalent, but a dark red-brown and yellow (see e.g. M.I. xiv. 0074, Hate L) are also met with. Ornamentation of individual scales was effected by scraping small designs such as concentric rings, ellipses, figures resembling commas or an inverted S, etc., through successive overlying coats of lacquer, after the fashion of sgraffito work. la some sets of scales (e. g. M. I. ix. 003, Plate L) bronze rivets are found, intended for ornamentation only. Fastening of That the methods of fastening the scales must have varied in detail in different suits Is shown by the diverse number and position of the holes to be found In certain sets of scales. Yet a comparison of those sets which still retain a number of scales as originally laced together (M.i, ix. 002, 003 ; xxiv. 0040, Plate L) shows that the system was similar In essential points. For the purpose of lateral fastening the longer sides of the scales were placed so as to overlap, and were then laced closely together by means of thongs passed through sets of holes near the edges. The exact fashion of this lacing has been explained in the descriptive notes of the List below* The vertical attachment was effected in a cleverly-designed way by two thongs running parallel through pairs of holes and round a thong which passed horizontally behind the rows of scales without ever appearing in front Owrfappiig It is certain that the rows of scales overlapped, and there is a probability that this overlapping &" was arranged upwards, in a fashion which differs from the classical and mediaeval examples of Europe but is illustrated by specimens of scale armour reproduced in Central-Asian and Graeco- Buddhist art, as well as by the Niya Site scales and the old Tibetan suit of mail already referred to.3 It is true that the scales to which I refer in these specimens, excepting the last named, are not oblong but of the round-headed type, and that where the lower or skirt portion of the same armour lias oblong scales, these are represented without any overlap, whether upwards or downwards. But It be noted that in the Tibetan suit of mail from Lhassa the scales are oblong throughout, overlapping upwards, and further that wherever we meet with scale armour in the Buddhist paintings tibe * Thousand Baddhas \ usually in figures representing the Guardian Kings of the quarters or die p0rtk» of the mail below the waist is made up of oblong scales always arranged * See p. 4%; I. p, 252; i. known relievo of the Lahore Museum, representing Ham's H H* bf two m a weft- anaj, see Foucber, £*art i$t GtmdMra^ L p. 405, %. 202.